---
license: llama2
datasets:
- wyt2000/InverseCoder-CL-13B-Evol-Instruct-90K
- ise-uiuc/Magicoder-Evol-Instruct-110K
library_name: transformers
pipeline_tag: text-generation
tags:
- code
model-index:
- name: InverseCoder-CL-13B
  results:
  - task:
      type: text-generation
    dataset:
      type: openai_humaneval
      name: HumanEval
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.799
      verified: false
  - task:
      type: text-generation
    dataset:
      type: openai_humaneval
      name: HumanEval(+)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.744
      verified: false
  - task:
      type: text-generation
    dataset:
      type: mbpp
      name: MBPP
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.746
      verified: false
  - task:
      type: text-generation
    dataset:
      type: mbpp
      name: MBPP(+)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.630
      verified: false
  - task:
      type: text-generation
    dataset:
      type: ds1000
      name: DS-1000 (Overall Completion)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.431
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Java)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.545
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (JavaScript)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.654
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (C++)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.581
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (PHP)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.553
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Swift)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.525
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Rust)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.556
      verified: false
  - task:
      type: text-generation
    dataset:
      type: nuprl/MultiPL-E
      name: MultiPL-HumanEval (Average for non-python languages)
    metrics:
    - name: pass@1
      type: pass@1
      value: 0.569
      verified: false
---
<div align="center">
  <img src="https://huggingface.co/wyt2000/InverseCoder-CL-7B/resolve/main/assets/logo.png" style="zoom:25%;" /> 
</div>

# InverseCoder: Unleashing the Power of Instruction-Tuned Code LLMs with Inverse-Instruct

<img src="https://huggingface.co/wyt2000/InverseCoder-CL-7B/resolve/main/assets/overview.png" style="zoom:50%;" /> 

InverseCoder is a series of code LLMs instruction-tuned by generating data from itself through Inverse-Instruct.

## Models and Datasets
|     | Base Model                                                                                           | InverseCoder                                                                                      | Dataset                                                                                                                              |
| --- | ---------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ |
| 6.7B | [deepseek-ai/deepseek-coder-6.7b-base](https://huggingface.co/deepseek-ai/deepseek-coder-6.7b-base) | [wyt2000/InverseCoder-DS-6.7B](https://huggingface.co/wyt2000/InverseCoder-DS-6.7B)               | [wyt2000/InverseCoder-DS-6.7B-Evol-Instruct-90K](https://huggingface.co/datasets/wyt2000/InverseCoder-DS-6.7B-Evol-Instruct-90K)     |
| 7B  | [codellama/CodeLlama-7b-Python-hf](https://huggingface.co/codellama/CodeLlama-7b-Python-hf)          | [wyt2000/InverseCoder-CL-7B](https://huggingface.co/wyt2000/InverseCoder-CL-7B)   | [wyt2000/InverseCoder-CL-7B-Evol-Instruct-90K](https://huggingface.co/datasets/wyt2000/InverseCoder-CL-7B-Evol-Instruct-90K)       |
| 13B  | [codellama/CodeLlama-13b-Python-hf](https://huggingface.co/codellama/CodeLlama-13b-Python-hf)          | [wyt2000/InverseCoder-CL-13B](https://huggingface.co/wyt2000/InverseCoder-CL-13B) **<= You are here** | [wyt2000/InverseCoder-CL-13B-Evol-Instruct-90K](https://huggingface.co/datasets/wyt2000/InverseCoder-CL-13B-Evol-Instruct-90K)       |

## Usage

Similar to [Magicoder-S-DS-6.7B](https://huggingface.co/ise-uiuc/Magicoder-S-DS-6.7B/), use the code below to get started with the model. Make sure you installed the [transformers](https://huggingface.co/docs/transformers/index) library.

```python
from transformers import pipeline
import torch
INVERSECODER_PROMPT = """You are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.
@@ Instruction
{instruction}
@@ Response
"""
instruction = <Your code instruction here>
prompt = INVERSECODER_PROMPT.format(instruction=instruction)
generator = pipeline(
    model="wyt2000/InverseCoder-CL-13B",
    task="text-generation",
    torch_dtype=torch.bfloat16,
    device_map="auto",
)
result = generator(prompt, max_length=1024, num_return_sequences=1, temperature=0.0)
print(result[0]["generated_text"])
```

## Paper
**Arxiv:** <https://arxiv.org/abs/2407.05700>

Please cite the paper if you use the models or datasets from InverseCoder.

```
@misc{wu2024inversecoderunleashingpowerinstructiontuned,
      title={InverseCoder: Unleashing the Power of Instruction-Tuned Code LLMs with Inverse-Instruct}, 
      author={Yutong Wu and Di Huang and Wenxuan Shi and Wei Wang and Lingzhe Gao and Shihao Liu and Ziyuan Nan and Kaizhao Yuan and Rui Zhang and Xishan Zhang and Zidong Du and Qi Guo and Yewen Pu and Dawei Yin and Xing Hu and Yunji Chen},
      year={2024},
      eprint={2407.05700},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2407.05700}, 
}
```

## Code

[Official code repo](https://github.com/wyt2000/InverseCoder) for Inverse-Instruct (under development).

## Acknowledgements

* [Magicoder](https://github.com/ise-uiuc/magicoder): Training code, original datasets and data decontamination
* [DeepSeek-Coder](https://github.com/deepseek-ai/DeepSeek-Coder): Base model for InverseCoder-DS
* [CodeLlama](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/): Base model for InverseCoder-CL
* [AutoMathText](https://github.com/yifanzhang-pro/AutoMathText): Self-evaluation and data selection method